Knowmind
Give your coding agent durable recall over projects—store facts, traverse relations, and check health on a hosted knowledge graph instead of re-explaining context every session.
Overview
Knowmind is a MCP server for the Build phase that provides knowledge graph storage, recall, relations, stats, and health checks as long-term memory for AI agents.
What is this MCP server?
- Knowledge graph and long-term memory: recall, relations, store, stats, and health tools
- stdio npm knowmind package v0.1.14 with positional mcp argument
- Required KNOWMIND_TOKEN from knowmind.de dashboard API tokens
- Configurable KNOWMIND_API_URL default https://knowmind.de
- Hosted service at knowmind.de with GitHub source WhoSoEverM8/knowmind
- Server package version 0.1.14
- MCP tools scope: recall, relations, store, stats, health per description
- Default API base https://knowmind.de
What problem does it solve?
Your agent forgets project context between sessions and you lack a structured graph to store and relate what the agent learned.
Who is it for?
Solo builders running multi-session agent projects who want hosted graph memory without self-hosting a full RAG stack.
Skip if: Teams that need only ephemeral chat context or already standardized on a single vendor memory with no MCP bridge.
What do I get? / Deliverables
Your agent reads and writes Knowmind memory through MCP so decisions, entities, and relationships persist across workflows with observable stats and health.
- Stored and recalled knowledge graph entries with relation queries
- Memory stats and health visibility for the connected agent workspace
Recommended MCP Servers
Journey fit
Agent-tooling in build is where you wire memory, tools, and persistence so the product and the agent stack evolve together. Long-term memory and graph relations are core agent infrastructure, not a one-off ship checklist item.
How it compares
Hosted knowledge-graph memory MCP, not a local markdown notes skill or generic web search tool.
Common Questions / FAQ
Who is Knowmind for?
Developers using MCP agents who need persistent knowledge graphs, recall, and relations across long-running indie product work.
When should I use Knowmind?
Use it during build and agent-tooling setup when you want store/recall/relation tools wired in, and keep using it as you operate and iterate on captured learnings.
How do I add Knowmind to my agent?
Install npm package Knowmind with argument mcp, set required KNOWMIND_TOKEN from https://Knowmind.de/dashboard/api-tokens, optionally set KNOWMIND_API_URL, and add the stdio server to your MCP config.